Review of GRDDL Primer

Here's my review of the Primer document:
http://research.talis.com/2006/grddl-wg/primer

Overall, this document is clear and informative. It is ready for
publication as a first working draft (modulo some small typos documented
below.) What follows is only suggestions.

If we really mean this to be a Primer, this document may be too much too
quickly. In a later revision, we may want to consider presenting fewer
use cases with more detail. A complete super-simple first example, with
one or two extracted triples at the most, e.g. dc:creator and dc:date,
the required XSLT and resulting output in RDF/XML. This could be the
beginning of the guitar example, or even simpler.

Once a simple example is fully presented, the more complex scheduling
example, in both MF and eRDF, are quite good as is.

In Status: "Publication as a does not imply ..."
missing "draft" or something else between "a" and "does"

Introduction could be a bit more brief. The discussion of XSLT 1.0 vs.
other methods seems out of place here.

Introduction, 3rd paragraph: "...to allowing a non-native speaker learn
the spoken..."
missing "to" before "learn."

Second Example, eRDF with scheduling:
should the CLASS of the P elements be "cal-Vevent" rather than
"-cal-Vevent"(note the extra "-")?

Example 3, "snippit" --> "snippet"


-Ben

Received on Tuesday, 26 September 2006 04:13:31 UTC